Background
Molten salt mixtures were imagined for use in nuclear reactors by Eugene Wigner during the Manhattan Project. Successful use of uranium hexafluoride in the K-25 gaseous diffusion uranium enrichment facility near Oak Ridge, Tennessee, built confidence in the use of uranium in fluoride form, and in 1950 a mixture of fluoride salts in liquid form was proposed to solve some of the issues associated with the Aircraft Nuclear Program. A small, proof-of-principle liquid-fluoride reactor was built and operated in 1954 at Oak Ridge, and two years later under the encouragement of laboratory director Alvin Weinberg, a more significant examination began of liquid-fluoride reactors for electrical generation at terrestrial power stations. Weinberg also encouraged the examination of the thorium fuel cycle implemented in liquid fluoride reactors, and this work led to the construction and operation of the Molten-Salt Reactor Experiment (MSRE) at Oak Ridge. The MSRE operated from 1965 to 1969, when it was shut down under the orders of Milton Shaw of the Atomic Energy Commission so as to free up additional funding for the liquid-metal fast breeder reactor (LMFBR) program. The molten-salt program continued for another three years at Oak Ridge until it was cancelled in 1972 under Shaw’s orders.
Introduction
Fluoride salt mixtures have many attractive features that recommend them for use in a nuclear reactor. They are very chemically stable and impervious to radiation damage due to their ionic bonding. Although they do not melt until operated at elevated temperatures (>350C) they have a wide liquid range beyond their melting point, approximately a thousand degrees Celsius. They dissolve useful quantities of actinide fluorides such as uranium tetrafluoride, thorium tetrafluoride, and plutonium trifluoride. They chemically capture fission products such as cesium and strontium in fluoride form and prevent their release. Most importantly, they operate at high temperatures yet at essentially ambient pressures, removing concerns about pressurized reactor operation.
Fluoride salt mixtures also have excellent volumetric heat capacity, somewhat better even than water. The volumetric heat capacity of the coolant is the basic yardstick that sizes a reactor and the rate at which its coolant must be pumped, giving fluoride salt reactors a great advantage over other designs.
Recently, the Department of Energy has funded work on a reactor concept called a fluoride high-temperature reactor (FHR) which uses solid uranium dioxide fuel (typically in TRISO-coated particles in a graphite matrix) cooled by a fluoride salt mixture (typically lithium fluoride and beryllium fluoride) to achieve high operating temperatures. Interest in liquid-fluoride reactors with dissolved nuclear fuel is also present at the industrial level with the formation of Flibe Energy in 2011. Flibe Energy is pursuing a design called a liquid-fluoride thorium reactor (LFTR), which is a modern variant of the work initiated at Oak Ridge during their research into molten-salt reactors. Other companies have also formed since 2011 to pursue molten-salt reactors, but their designs have not incorporated the thorium fuel cycle, relying instead on enriched uranium.
The LFTR concept has attracted the attention of regulated utilities that have funded research work through the Electric Power Research Institute (EPRI) to help further define LFTR subsystems. Regulated utilities build power plants based on the consent of state public service commissions (PSC) that seek the lowest levelized cost-of-electricity (LCOE) possible for their ratepayers. Since regulated utilities earn profit on the capital deployed in the construction of power plants, it is desirable to pursue a technological solution that has a competitive LCOE while minimizing costs that do not earn profit for the utility, such as fuel costs and operations and maintenance (O&M) costs. For natural-gas-fired turbines, fuel costs are a dominant term in the assessment of LCOE, but for nuclear power plants fuel costs are a much smaller fraction of LCOE costs. At current natural gas prices it is difficult for conventional nuclear power concepts to compete with natural-gas-fired turbines on the basis of lowest LCOE, therefore it is highly desirable from a utility’s perspective to develop a new form of nuclear energy that can be competitive against natural gas on this basis. Since natural gas prices fluctuate substantially over time, it is also in the best interest of PSCs to minimize the risk to the ratepayers of having too much natural gas deployed in a particular region. Regulated utilities do not earn profit on fuel for their coal-fired or gas-fired facilities, and there is no advantage to a regulated utility if gas prices go up, since these costs are passed on directly to the consumer.
Fuel Cycle
In today’s uranium reactors, natural uranium is mined, purified, and chemically converted to uranium hexafluoride prior to enrichment. The process of enrichment results in two output streams, one enriched and one depleted. Approximately five parts out of six of the original uranium ends up in the depleted stream, with only one part out of six going on to be fabricated into nuclear fuel. Most of the small amount (0.7%) of the original uranium that is uranium-235 ends up in the enriched stream, but about a third remains in the depleted stream. The enriched uranium hexafluoride is chemically converted back to uranium dioxide, pressed into pellets, loaded into zirconium tubes to form fuel rods, and arranged into clusters to form fuel assemblies. These assemblies are then loaded into a reactor where they will spend approximately five years in the core in various locations to generate nuclear energy. At the end of their useful life, they are removed from the reactor and allowed to cool in a spent fuel pond. Most of their uranium-235 has been consumed. Some of their uranium-238 has been converted into plutonium-239 and some of that has also been consumed. Fission products have been generated from the fission of both uranium-235 and plutonium-239. Several long-lived isotopes of plutonium, americium, and curium have been formed when plutonium-239 absorbed a neutron rather than fissioned. These long-lived actinides present a disposal challenge, yet their formation at a rate that exceeds their consumption is inevitable when uranium fuel is used in a thermal-spectrum reactor of any type, including a molten-salt reactor.
In a future liquid-fluoride thorium reactor, the fuel cycle would be quite different. The reactor would ideally be started by a modest inventory of uranium-233. Fission of U-233 in the reactor generates thermal power as well as excess neutrons that would be captured in a blanket fluid containing thorium tetrafluoride in solution. Thorium, having absorbed a neutron, first decays to protactinium and ultimately to uranium-233. New fuel would be chemically removed from the blanket fluid either at the uranium stage or the protactinium stage, which has additional complexity and advantages. The new uranium fuel would be introduced into the fuel salt of the LFTR at the same rate at which it is consumed. Uranium-233 is consumed at high efficiency (91%) in a thermal spectrum reactor and that which is not consumed goes on to form uranium-235, which is also consumed at high efficiency (85%) in a thermal spectrum reactor. This limits considerably the amount of material that can reach the stage of the first transuranic, in this case, neptunium-237, and thus the issue of long-lived actinide waste production. The fuel salt used in the LFTR is chemically processed as the reactor operates, removing fission products while retaining actinide fuels. This allows the creation of an actinide-free waste stream which decays to acceptable radioactivity levels in approximately 300 years, strongly governed by the 30-year half-lives of cesium-137 and strontium-90.
Energy Markets
As previously mentioned, regulated investor-owned utilities generate profit for their shareholders through the guaranteed return they earn on the construction of new facilities. Hence, they are in a continuous state of examining ways in which they can expand and modernize their generation fleet. The public exercises its opinion through the state public service commissions, which give voters and ratepayers a say in the operation of these utilities. It is in the best interest of both parties to build and operate efficient power-generation facilities that minimize the cost to the consumer through the efficient use of fuel, no matter what type of fuel that is.
Thorium fuel used in a LFTR has the potential to achieve approximately 200 times the fuel efficiency as uranium fuel used in existing reactors. In addition, the LFTR operates at high temperatures and this makes possible the implementation of power conversion systems based on supercritical carbon dioxide gas that can achieve thermal efficiencies of approximately 45%, meaning less of the thermal power generated by the reactor need be rejected to the environment, and more can be converted to electrical energy for the consumer.
The high operating temperatures of the LFTR also enable direct applications of its process heat to be considered, such as the thermochemical generation of hydrogen, which could become an important part of the transportation infrastructure of the country in the future.
Nuclear fission also produces other valuable materials as the products of fission, most notably several important medical radioisotopes. Each of these products is characterized by a rather short half-life, which means that existing solid-fueled reactors cannot extract them quickly enough before their value is lost to decay. Currently these isotopes are produced in dedicated materials testing reactors in Canada and the Netherlands. Unfortunately each of these reactors is scheduled to be permanently shut down next year, putting world supply at risk. The LFTR not only produces these isotopes but its fluid fuel means that they can be readily extracted. This can stabilize world supply and provide an additional source of revenue for the reactor’s operator.
Through the realization of additional revenue through products like hydrogen, process heat, medical radioisotopes, and even desalinated seawater, the levelized cost of electricity that would need to be charged by a reactor’s operator could potentially be reduced, making LFTR construction and deployment feasible even with the prospect of low-cost natural gas.
Safety
Today’s nuclear reactors have an excellent safety record but are nevertheless highly feared by many members of the public, despite the fact that no one has been killed in the United States from the operation of civilian power reactors. Why is there such a huge dichotomy between perception and reality? One of the reasons why is the fear of radioactive exposure from the dispersal of radionuclides following a severe accident. Such a situation was realized in Japan after the terrible Tohuku earthquake of March 2011 and the tsunami it spawned. Although all the reactors struck by the tsunami survived, at the Fukushima-Daiichi plant the emergency core cooling system was damaged due to the loss of diesel-electric power generators. Decay heat from spent fuel compromised the integrity of several of the reactors, leading to zirconium-water reactions that produced hydrogen gas, which was vented from the containment and detonated spectacularly in other parts of the reactor building. Radionuclides, notably cesium from damaged fuel, spread across the surrounding area and is the primary contributor to elevated levels of radioactive exposure that persist to this day.
Although no one was killed by the events at Fukushima-Daiichi, it is clear that the nuclear industry must do better if public support of nuclear energy is to continue. The General Electric boiling water reactor Mark I that was used at Fukushima-Daiichi is a significant part of the US reactor fleet, with 22 reactors of this design in operation.
The public fears exposure from radionuclides that could be dispersed from a severe accident. The reactor designer should consider ways to reduce the potential for the dispersal of such materials. The LFTR design gives many advantages in such an effort. The reactor is unpressurized, removing pressure as a driving term for radionuclide dispersal. Decay heat is handled through a passive system where fluid fuel drains into a dedicated tank, and this drain is mediated by a “freeze valve”, a frozen plug of salt that is actively cooled to keep it in place. Without active cooling, the frozen plug melts and the salt drains into a passively-cooled configuration. The fluoride chemical form of the fuel is another important advantage, chemically trapping fission products such as cesium, strontium, and iodine in chemically stable forms and strongly limiting their potential for dispersal. The fact that the salt is only fluid at high temperatures is another advantage for reactor safety, meaning that if it is cooled to ambient temperatures it freezes and traps fission products therein.
Challenges
There are significant challenges that remain before LFTRs can be deployed in the scale necessary. The high-nickel alloy proposed for use with the reactor (Hastelloy-N) must be ASME code-qualified to permit certified construction. Suitable core designs that achieve safety and neutronic goals must be refined and tested. The carbon dioxide power conversion system must be scaled up and demonstrated. Capture of tritium formed from neutron absorption in lithium salt must be demonstrated to satisfaction. Suitable heat exchanger designs for a variety of locations in the reactor must be completed, tested, and proven. The chemical processing system that enables the thorium fuel cycle must be demonstrated at a prototype scale, and then at a larger bench scale, before ultimately being tested in a real reactor. Associated with this will be the need to demonstrate that long-lived actinides can be excluded from the waste stream through the proper operation of the chemical processing system. And perhaps most importantly, a business case that accounts for the totality of potential products from the reactor must be developed and shown to be competitive with low-cost hydrocarbon fuels.
Sustainability
Assuming that these challenges can be met, the thorium fuel cycle implemented in the LFTR promises to have exceptional sustainability. Producing a gigawatt of electricity for a year would only consume less than a tonne of thorium fuel, and the United States has 3200 metric tonnes of thorium in easily accessible disposal areas, in addition to hundreds of thousands of tonnes in geologic deposits like the Lemhi Pass area of Idaho. If used efficiently in a LFTR, thorium could provide energy security for the United States for the foreseeable future and likely beyond.
Conclusion
The liquid-fluoride thorium reactor concept has strong safety advantages over today’s nuclear reactors and the potential to implement a highly efficient and sustainable fuel cycle. It can potentially produce valuable products in addition to electrical energy that will enhance its competitiveness relative to low-cost natural gas and petroleum. The capital cost structure of a LFTR would make it attractive to regulated electrical utilities that desire to maximize shareholder return while providing low-cost electrical energy to the ratepayers in their service area.